Zero-Shot Voice Cloning & Paralinguistic Control
Target Text to Synthesize
Click to insert paralinguistic tag (Chinese & English):
Target Language
English
Chinese (中文)
Japanese (日本語)
Korean (한국어)
German (Deutsch)
French (Français)
Spanish (Español)
Preset Reference Clip
Preset: Lingguang (Warm Conversational)
Preset: DB30 (Clear Studio Speaker)
Custom Uploaded Audio File
One-Click Examples
🇨🇳 傍晚的风从山口吹过来，远处的云慢慢染成了金色。
🎭 [Sigh]雨还是没有停，[Cough]我们恐怕只能把活动改到室内了。
🇬🇧 The old radio crackled softly as the midnight storm moved across the valley.
🇯🇵 週末、駅前の小さな喫茶店で、季節限定の抹茶パフェを食べました。
Text-to-Voice Design (Zero Reference Audio Needed)
Text to Speak
Speaker & Timbre Description
One-Click Examples
🎬 成熟沉稳的男性纪录片旁白 — 冰川在漫长岁月中缓缓移动…
✈️ British Female Airline Captain — Welcome aboard Flight 702 to Tokyo…
117 Built-in Voices (5 English + 112 Chinese Characters)
Text to Speak
Emotional Variant
One-Click Examples
🇬🇧 UK_Charlotte — The botanical garden will remain open until sunset today.
🇺🇸 US_James — Good evening everyone, and welcome back to our live broadcast!
🇨🇳 灵阅川 — 山谷里的雾气还没有散开，远处的钟声已经轻轻传了过来。
Explicit Emotion & Prosodic Attribute Control
Text to Speak
Control Mode
Explicit Emotion
Speed, Pitch & Volume
Emotion
Happy (高兴)
Surprised (惊讶)
Sad (悲伤)
Angry (愤怒)
Fearful (恐惧)
Disgusted (厌恶)
Neutral (中性)
Speed
Slow (慢速) Medium (中速) Fast (快速)
Pitch
Low (低) Medium (中) High (高)
Volume
Low (低) Medium (中) High (高)
Cantonese (广粤话) & Sichuanese (川渝话) Dialect Synthesis
Dialect Text
Target Dialect
Cantonese (广粤话)
Sichuanese (川渝话)
One-Click Examples
🇭🇰 Cantonese (广粤话) — 今日天气几好，我哋食完饭一齐去海边行下啦。
🌶️ Sichuanese (川渝话) — 今天天气巴适得很，我们吃完火锅去江边走一圈嘛。
Two-Speaker Conversational Podcast (`podcast`)
Dialogue Script (Start every line with speaker_1: or speaker_2:)
⚡ Synthesize Speech with Ming-3.0-tts-tiny